Isolating local, background, and meteorological contributors to urban air pollution using wavelet decomposition of multi-pollutant sensor network data in Los Angeles, CA
Journal Article
Overview
abstract
Abstract; ; A multi-pollutant sensor network measuring methane (CH; ; ; ; ; ; ; ; ; 4; ; ; ; ; ), carbon monoxide (CO), nitrogen dioxide (NO; ; ; ; ; ; ; ; ; 2; ; ; ; ; ), particulate matter, volatile organic compounds (VOCs), and benzene, toluene, ethylbenzene, and xylenes (BTEXs) was deployed for nearly two years to investigate the sources and variability of urban air pollution in a Los Angeles community. To distinguish the processes driving pollution variability, we applied wavelet decomposition, a time–frequency analysis technique that separates high- and low-frequency components within air quality data. Overall pollution concentrations varied seasonally, with higher levels during colder months, but showed relatively small differences across monitoring locations within the network. Wavelet decomposition isolated contributions from high-frequency events, persistent background enhancements, and diurnal trends related to meteorology. The high-frequency component peaked during the morning rush hour and exhibited the greatest spatiotemporal heterogeneity across the network, indicating this component is largely influenced by local emissions. In contrast, the background component contributed substantially to overall pollution levels but showed much lower spatial variability. Meteorologically-driven pollution was highest at night, when low wind speeds and reduced boundary layer heights limited pollutant dispersion. This diurnal component was also associated with temperature-driven VOC and BTEX formation and reductions in NO; ; ; ; ; ; ; ; ; 2; ; ; ; ; . These results demonstrate how wavelet decomposition can effectively distinguish drivers of urban air pollution pollution based on their varying temporal frequencies.;
